Skip to main content
URMC / Education / Graduate Education / URBest Blog / June 2018 / From PhD to Data Science

From PhD to Data Science

Career Story By Aslihan Ambeskovic, PhD, Head of Cancer Bioinformatics and Research Project Manager 

Before I started my career as a data scientist, what I kept hearing was how it was the "sexiest job of the 21st century". I always thought using the word "sexy" for non-sexy entities just meant that everybody wanted it but nobody knew why.

Once in the trenches it was easy to understand why being a data scientist was desirable by many and why almost every company that had an app/web presence needed one. From an individual point of view, the job keeps you fully fed, it never lacks intellectual stimulation, and interesting dinner topics such as "do you want to hear what my recurrent neural network trained on (insert an ostentatious political figure name here)'s tweets came up with" are plenty. It is also easy to understand why many companies need data scientists. As the internet usage became as widespread as the need for education, it became feasible for companies to collect a multitude of data points from their users, passively and actively. A company can turn those data points into new products, customers, revenue and decisions with the help of data scientists.

If you are in a data intensive research area, maybe data science is not a big leap for your next step in your career. As the line between business and science blurs, PhDs are highly sought after as data scientists for their critical thinking and data skills. However, you need to master some other skills mentioned below as well to pass a technical interview for the position.

-Coding: From data exploration, visualization to modeling, multiple steps of a data science project involve coding. The language of choice is Python, followed by R for data science.

-Machine Learning: It is crucial for a data scientist to understand the statistical properties of the data and the problem at hand to be able to select the appropriate solution for it.

-Data Wrangling: The data in the real world, in contrast to the PhD world, is never clean. Get ready to spend a good chunk of your time data cleaning if you choose to be a data scientist.

-Communication:  If you can't communicate your results to your team, then your results don't exist. It is not as harsh as the publish or perish mentality of the sciences but it is key to be able to talk about to your results to a technical and a non-technical audience.

The best way to master all these skills is by practicing them in a project. There are many boot camps out there that would allow you to practice these skills and help you connect with companies who need data scientists. Also, keep in mind that every company and their data problems are different. If you are planning on moving from studying protein interactions to recommending art work for art lovers, make sure that you do some background research about the domain that you would like to move into.

I love data and data science. I recently came back to the university as a cancer bioinformatician to apply my data science skills to cancer ‘omics. If you would like to learn more about data science or if you are planning on becoming one, join me on June 5th at 3:30 PM in the Medical Center, 1-9525 Northeastern Conference Room.

Tracey Baas | 5/31/2018

You may also like

No related posts found.